Comparison of the Functional Properties of the Monomeric and Dimeric Forms of the Isolated CP47-Reaction Center Complex*

Chlorophyll fluorescence, thermoluminescence, and EPR spectroscopy have been used to investigate the functional properties of the monomeric and dimeric forms of the photosystem II CP47-reaction center (CP47-RC) subcore complex that was isolated (Zheleva, D., Sharma, J., Panico, M., Morris, H. R., and Barber, J. (1998) J. Biol. Chem. 273, 16122–16127). Chlorophyll fluorescence yield changes induced either by the initiation of continuous actinic light or by repetitive light flashes indicated that the dimeric, but not the monomeric, form of the CP47-RC complex showed secondary electron transport properties indicative of QA reduction. Thermoluminescence measurements also clearly distinguished the monomer from the dimer in that the latter showed a ZV band, which appeared at −55 °C, following illumination at −80 °C. This band has been determined to be an indicator of the photoaccumulation of QA·̄. The ability of the dimeric CP47-RC to show secondary electron transport properties was clearly demonstrated by EPR studies. The dimer was characterized by organic radical signals at about g = 2 induced either by illumination or by the addition of dithionite. The dithionite-induced signal was attributed to QA·̄, but there was no indication of any interaction with non-heme iron. The signal induced by light was more complex, being composed not only of the QA·̄ radical but also of radicals generated on the donor side. Difference analyses indicated that one of these radicals is likely to be due to a D1 tyrosine 161 or D2 tyrosine 161. In contrast, the monomeric CP47-RC complex did not show similar EPR-detectable radicals and instead was dominated by a high yield of the spin-polarized triplet signal generated by recombination reactions between the oxidized primary reductant, pheophytin, and the primary donor, P680. It is also concluded from EPR analyses that both the monomeric and dimeric forms of the CP47-RC subcore complex contain one cytochrome b 559 per reaction center. Overall the results suggest that photosystem II normally functions as a dimer complex and that monomerization at the level of the CP47-RC subcore complex leads to destabilization of the bound plastoquinone, which functions as QA.

at the level of the CP47-RC subcore complex leads to destabilization of the bound plastoquinone, which functions as Q A .
More than 25 different protein subunits make up the photosystem II (PSII) 1 complex of oxygenic photosynthetic organisms (1). At the heart of this complex is the reaction center (RC) consisting of the D1 and D2 proteins, where primary charge separation occurs (2). Closely associated with the D1 and D2 proteins are two similar chlorophyll a-binding proteins, CP43 and CP47 (3). These proteins serve as an "inner antennae" system that is linked to a secondary light-harvesting system. In higher plants and green algae, the chlorophyll a/bbinding proteins (encoded by nuclear located cab genes) act as the secondary light-harvesting system, while phycobilisomes serve the same purpose in other types of oxygenic photosynthetic organisms, such as red algae and cyanobacteria (4). CP43 and CP47 are also distinguished by having a large hydrophilic loop linking putative membrane-spanning regions 5 and 6 (3). These loops are almost certainly located on the luminal surface of the complex and may function in water splitting in some way (5). Treatments with detergents can peel away the various subunits, and it has been shown that during such manipulations CP43 is more readily removed than CP47 (6,7). It is therefore possible to isolate a CP47-RC complex. Recently a method was described for spinach that yielded a preparation of the CP47-RC complex consisting of a mixture of monomeric and dimeric forms (8). Analyses using mass spectrometry showed that both forms of this subcore complex contained the products of the psbE, psbF, psbI, psbT c , and psbW genes as well as the D1 and D2 proteins and CP47. However, the CP47-RC dimer contained, in addition, the products of the psbL and psbK genes. Also of significance was the finding that the dimer and not the monomer contained about one molecule of plastoquinone-9 per RC. Overall, the findings suggest that the monomeric form of the isolated CP47-RC complex was derived by dissociation of the dimer and that the latter conformation is likely to be the in vivo state. Moreover, the finding that the CP47-RC complexes contain several small proteins with putative one transmembrane helices is relevant to de-* This work was supported in part by the Biotechnology and Biological Research Council and Hungarian Granting Agency OTKA Grant T 017049 (to I. V.). Financial assistance from The Royal Society and British Council allowed collaboration between J. B. and I. V. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  tailed structural studies that are under way on the complex using electron crystallography (8,9).
In this paper, we characterize the functional properties of the CP47-RC complex isolated from spinach and, in particular, identify differences between the monomeric and dimeric forms. We also address the question of the stoichiometry of cytochrome b 559 in these preparations.

MATERIALS AND METHODS
Isolation of Monomeric and Dimeric CP47-RC Complexes-The method of obtaining isolated CP47-RC in monomeric and dimeric forms from spinach thylakoids involves solubilization of PSII-enriched membranes using the detergents n-dodecyl ␤-D-maltoside and heptyl thioglycopyranoside, followed by Fractogel ion exchange chromatography and sucrose gradient centrifugation. The procedures have been given in detail previously (8).
Fluorescence Measurements-Relative chlorophyll fluorescence yields and time-dependent changes in the yield of chlorophyll fluorescence were measured using a PAM fluorimeter (WALZ, Effeltrich, Germany). Samples of 30 g of Chl ml Ϫ1 were illuminated with either continuous, red, actinic light (800 E m Ϫ2 s Ϫ1 ) or single turnover, saturating xenon flashes. The intensity of the modulated measuring light was 8 E m Ϫ2 s Ϫ1 , and the ms induction curves were analyzed using a QA Data acquisition software package (10). In the case of flash excitation, 20 fluorescence traces were averaged in order to increase the signal:noise ratio. Samples were dark-adapted before the measurements for 3 min, and for flash excitation the subsequent flashes were separated by 20-s dark intervals.
Thermoluminescence Measurements-Thermoluminescence curves were measured with a home-built apparatus as described earlier (11). Samples, containing 40 g of Chl, were dark-adapted for 3 min and then cooled to Ϫ80°C in darkness. Illumination with continuous white light of 10 watts m Ϫ2 was administered at Ϫ80°C for 30 s, followed by heating in the dark at a constant heating rate of 20°C min Ϫ1 .
EPR Spectrometry-For EPR, 0.3-0.4-ml samples (about 0.3 mg of Chl) were placed in calibrated 3-mm quartz EPR tubes. Dim green lighting was used to minimize photodamage to samples. Identical sets of samples in calibrated EPR tubes were made for each experiment. Samples were dark-adapted for a minimum of 30 min on ice before being frozen to 77 K in the dark. For chemically reduced samples, the addition of dithionite (25 l/sample) was made from a freshly made concentrated stock solution (1 g/50 ml) in 100 mM Tris/HCl buffer bubbled with oxygen-free nitrogen. During experiments, samples were illuminated at a variety of temperatures from 4 to 273 K as indicated. Illumination in the EPR cavity at Ͻ30 K was accomplished using a 150-watt light source and fiber-optic light guide, while other illumination was from a 1000-watt light source, protecting the sample from heating where necessary by a 5-cm water filter. Samples for freezing under illumination to 200 K were made using an ethanol/dry ice bath in a clear glass Dewar. The samples were then quickly transferred to liquid nitrogen in the dark.
Samples were examined by EPR at cryogenic temperatures using a Jeol RE1X X-band spectrometer (9.055 GHz) with 100-kHz modulation and fitted with an Oxford Instruments cryostat. EPR conditions are given in the figure legends. Temperatures within the EPR cryostat were measured with a calibrated thermocouple beneath the sample. Spectra were recorded and manipulated using a Dell microcomputer running Asyst software. No filtering, smoothing, fitting, or background subtractions were used. The vertical scale in figures showing first derivative EPR spectra is arbitrary, with spectra at the same instrument gain unless stated in the figure legend.
The cytochrome b 559 content of CP47-RC monomers and dimers was measured by reference to a sample of the D1-D2-cytochrome b559 RC complex (5.9 Chl/2 Pheo, 242 g of Chl ml Ϫ1 ). It is assumed that this preparation had one cytochrome b 559 per reaction center (12). The integrated area of the g z peak of this sample was compared with that of two preparations each of PSII monomers (17 and 18.2 Chl/2 Pheo) and dimers (20.8 and 23 Chl/2 Pheo). Accurate measurement of both the chlorophyll concentration and the Chl/Pheo ratios then enabled the ratio of cytochrome b 559 in the PSII monomers and dimers to be calculated. The chlorophyll levels were measured as described previously (8,13).

RESULTS
Chlorophyll Fluorescence-CP47-RC monomers and dimers were analyzed by measuring variable chlorophyll fluorescence yield to detect their electron transfer activities at room temperature. Fig. 1 shows photoinduced changes of PSII chlorophyll fluorescence yield (⌬F) of dimeric and monomeric CP47-RC subcore complexes. The results clearly show that the quinone within the dimeric form of the complex can act as a secondary electron acceptor indicative of Q A activity and because of this exhibits fluorescence characteristics distinctly different from those of the monomeric form of the complex. Saturating actinic light in the absence of any additions induced an increase of PSII chlorophyll fluorescence yield (⌬F) of dimeric CP47-RC ( Fig. 1, curve 1). The ⌬F/F o (F o , so called "constant" F) ratio was about 3, which is similar to the ratio of PSII-enriched membranes (BBYs). The magnitude of ⌬F did not increase in the presence of PSII artificial electron donors (data not shown). The real magnitude of F o was confirmed using K-15, a compound that selectively quenches the part of the PSII chlorophyll emission that is due to recombinant luminescence (14). When the actinic light was turned off, the dark relaxation of the ⌬F (reflecting the reoxidation of the primary electron acceptor of PSII, Q A Ϫ ) was significantly slower for the dimeric CP47-RC compared with that of BBYs ( Fig. 1, curve 1, dashed line); less than 20% of the maximal level ⌬F is quenched after 150 s in the dark. However, repeated measurements of ⌬F on the same CP47-RC sample (after achieving complete reoxidation of Q A Ϫ in the dark) resulted in a lowering of ⌬F. Subsequent additions of artificial electron donors of PSII, Mn 2ϩ (0.1-20 M), sodium ascorbate (2 mM), diphenylcarbazide (1 mM), and NH 2 OH (1 mM) to the sample did not restore the initial level of photochemical activity (data not shown). Fig.  1, curve 2, shows the photoinduced chlorophyll fluorescence yield of the dimeric CP47-RC subcore complex measured in the presence of 1 mg ml Ϫ1 sodium dithionite, which chemically reduces Q A in the dark. In this case, the fluorescence rose to its maximal level (F m ) when the sample was illuminated only with measuring light, whereas the application of saturating actinic light caused a reversible decrease of F m (to the level F o R ), related to the photoreduction of pheophytin, as shown earlier (15). The fluorescence behavior of the monomeric CP47-RC subcore complex differed significantly from dimeric CP47-RC as well as from PSII-enriched membranes and was similar to that of D1-D2-cytochrome b 559 RC complex. In the absence of actinic light and any additions, switching on the weak measuring light induced an increase of F to its maximal level, F m ( Fig. 1, curve 3), as is also observed for the D1-D2-cytochrome b 559 RC complex (Fig. 1, curve 5). No positive increase of PSII chlorophyll fluorescence yield was observed when the saturation actinic light was switched on. However, in contrast to the RC preparations, where a short illumination (a few tens of seconds) with saturating actinic light did not induce any significant quenching of F m in the absence of sodium dithionite ( Fig. 1, curve 5), in the monomeric CP47-RC it did induce a significant quenching of F m (Fig. 1, curve 3). In the presence of sodium dithionite, however, the magnitudes of F m quenching isolated RC and CP47-RC complexes were comparable ( Fig. 1, curves 4 and 6). The addition of a range of PSII artificial electron donors did not induce a ⌬F response (not shown). The addition of K-15 to the monomeric CP47-RC preparation caused quenching of F to the F o level of the dimeric form of the complexes (data not shown). In the presence of sodium dithionite, illumination of the monomeric CP47-RC by actinic light resulted in fast quenching of F m to the level of F o R (Fig. 1, curve  4). The time course of the quenching coincides with those recorded for both the dimeric CP47-RC and RC complexes, and the F o R level corresponded to that obtained in the presence of K-15. Also of note is that the light-induced ⌬F quenching in the presence of dithionite for all samples is reversible when the actinic light is turned off (Fig. 1, curves 2, 4, and 6).
The capability of CP47-RC dimer preparation to reduce Q A was probed further by measuring the kinetics of chlorophyll fluorescence yield changes when induced by single turnover saturating flashes. As shown in Fig. 2, there was no flashinduced fluorescence yield change in the monomers, as expected from the steady-state fluorescence measurements shown in Fig. 1. In the dimers, however, a flash-induced rise of fluorescence yield was clearly observed, indicating that Q A can be reduced by a single flash (Fig. 2, curve B). The subsequent decay of the fluorescence yield reflects reoxidation of the semiquinone anion radical Q A . , which is characterized by a few hundred-ms time constant in the dimers. This decay time is much slower than the 400 -500-s and 3-4-ms values assigned to reoxidation of Q A . by forward electron flow to Q B in fully functional PSII particles or BBYs (see Fig. 1) (16). In addition, the presence of DCMU, which blocks electron transfer between Q A and Q B , had no effect on the flash-induced fluorescence yield relaxation kinetics (not shown), indicating that there is no electron transfer to a secondary quinone in the Q B pocket in dimers. These results are in line with the conclusions drawn from the steady-state fluorescence studies and are consistent with the finding that the dimer, but not the monomer, of CP47-RC contains a plastoquinone molecule (8).
Thermoluminescence-Electron transfer properties of the CP47-RC monomers and dimers were further probed by measuring thermoluminescence signals (Fig. 3). A characteristic thermoluminescence component observed with PSII complexes with a small number of subunits is called a Z v band, which appears in the Ϫ80 to Ϫ30°C range, having peak temperatures that exceed the illumination temperature by about 20°C (17). It has been shown that the intensity of the Z v band indicates the presence of functional Q A , since it can be partially restored by the addition of quinones to quinone-free D1-D2-cytochrome b 559 RC complexes (18).
Thermoluminescence measurements performed with CP47-RC dimers show a Z v band that appears at Ϫ55°C, after excitation at Ϫ80°C. In the monomeric complexes, the Z v band is largely missing, and only a small thermoluminescence signal is observed with peaks at Ϫ55 and Ϫ30°C. The origin of the Ϫ30°C component is not clear, but this weak band is also present in D1-D2-cytochrome b 559 RC complexes (11) and may be hidden under the large Z v band of the CP47-RC dimers. The significant difference in the intensity of the Z v band in the two types of CP47-RC preparations is consistent with the conclusion that the monomers retain virtually no functional Q A compared with the dimers. More detailed analysis of the data reveals, however, that the area under the Ϫ55°C component in the monomers corresponds to about 20% of the Z v band in the dimers. If the integrated intensities of these bands reflect the amount of light reducible Q A , as expected, this would indicate about 0.2 active quinones/RC in the monomers. This value seems to be somewhat high, given the almost complete lack of light-inducible fluorescence yield change (Figs. 1 and 2) and the absence of detectable plastoquinone-9 in the monomeric sample (8). A possible explanation for this apparent discrepancy is that the Z v band may not directly originate from Q A . but is enhanced by its presence. This possibility is reinforced by the fact that a small Z V band has also been observed with the D1-D2cytochrome b 559 RC complex having no plastoquinone present (11). EPR Spectrometry-CP47-RC monomers and dimers were also analyzed by EPR to detect their electron transfer capabilities at cryogenic temperatures. Fig. 4 shows that the CP47-RC monomers gave a high yield of spin-polarized reaction center triplet (Fig. 4B) compared with PSII dimers (Fig. 4A). The high yield of spin polarized triplet observed with the monomers is reminiscent of that recorded previously with the isolated D1-D2-cytochrome b 559 RC complex (19). The CP47-RC dimers showed a maximum of 12.5% triplet yield compared with the monomers. The g ϭ 2 region of the EPR spectrum was examined for organic radicals. The formation of a radical by illumination showed that the CP47-RC dimers were capable of stable charge separation (Fig. 4C), while the monomers showed much less activity (Fig. 4D). No tyrosine radical was present, and no rapidly reversible light-induced signals were detected in either preparation. The light-induced g ϭ 2.003 signal (Fig. 4C) in CP47-RC dimers had characteristics suggesting it was a mixture of at least two radical species. Taken together, these results indicate that the CP47-RC monomers show a similar capability to the D1-D2-cytochrome b 559 RC complex with electron transfer at cryogenic temperatures restricted to the forward and back reactions between the primary donor P680 and primary acceptor pheophytin (19,20). However, CP47-RC dimers are capable of electron transfer beyond pheophytin, thereby reducing the triplet yield and producing the g ϭ 2 radical on illumination. Fig. 5, A and B, shows that low spin cytochrome b 559 was present in both CP47-RC monomers and dimers, the latter having a broader g z peak near 220 mT. In both preparations, the cytochrome b 559 was in the low potential, fully autoxidized form, and no light-induced changes were observed. As outlined under "Materials and Methods," the content of cytochrome b 559 was estimated by comparison with the D1-D2-cytochrome b 559 RC preparation, which is assumed to have one heme per reaction center (13). This comparison shows that CP47-RC monomer preparations (0.92 and 0.81 per RC) and dimers (0.97 and 0.90 per RC) had slightly less than one cytochrome b 559 per RC. An additional peak near g ϭ 6 was observed in both monomer and dimer preparations (not shown). This was larger in the preparations with lower ratios and may indicate that some conversion of low spin heme to high spin heme occurs. We conclude that there is one cytochrome b 559 per reaction center in both preparations.
The EPR results confirm the data obtained from chlorophyll fluorescence and thermoluminescence indicating that the additional electron transfer component is present in the dimeric form of the CP47-RC complex and that this component is likely to be a plastoquinone functioning as Q A . However, careful analysis of EPR radical spectra attributed to this component showed almost no formation of any type of iron-semiquinone EPR signal (21)(22)(23). The use of PSII membranes (i.e. BBY-type) of similar reaction center concentration was used to confirm that a Ͻ Ͻ10% reaction center concentration of iron-semiquinone would have been detected. A possible explanation was that the non-heme iron, normally located between the Q A and Q B sites in PSII, was absent in the CP47-RC dimer. Fig. 5 (right side) shows that in CP47-RC dimers following chemical reduction, an organic radical was detected at g ϭ 2.
The radical (Fig. 5C) had Hpp ϭ 0.95 mT, g ϭ 2.005 (where Hpp represents peak to trough line width of the EPR spectrum) and was microwave power-saturated above 1 microwatts. This is characteristic of an anionic semiquinone such as might be expected if Q A . were present but the non-heme iron was absent.
A similar signal was observed in PSII membranes treated with cyanide to remove the interaction between Q A . and the nonheme iron (24,25). By contrast, the PSII monomer sample shows no chemically induced radical under these conditions (Fig. 5D).
The results of Figs. 4 and 5 therefore show that CP47-RC dimers retain Q A and are capable of electron transfer to it, forming the semiquinone. The non-heme iron seems to have been lost, showing that this cofactor can be removed before Q A and before isolation of the D1-D2-cytochrome b 559 RC complex, where it is known to be absent (26). The mixture of radicals obtained in Fig. 4C can now be assigned to the Q A semiquinone and the chlorophyll/carotenoid usually found to be an electron donor at these temperatures (19,20). The presence of Q A as electron acceptor may allow other electron donors to function. No tyrosine radical was detected in dark-adapted samples of PSII monomers or dimers. However, by freezing a PSII dimer sample under illumination, a complex spectrum at g ϭ 2 was FIG. 4. EPR spectra showing the effects of illumination at cryogenic temperatures. Left, light-minus-dark difference spectra at 4.2 K showing the yield of spin-polarized reaction center triplet in CP47-RC dimers (1.002 mg of Chl/ ml) (A) and CP47-RC monomers (0.776 g of Chl/ml) (B). EPR conditions were as follows: microwave power, 40 microwatts; modulation amplitude, 2 mT. Right, spectra near g ϭ 2, taken dark and then dark after 2-min illumination at 8 K of CP47-RC dimers (C) and CP47-RC monomers (D). EPR conditions were as follows: microwave power, 1 microwatts; modulation amplitude, 0.2 mT. For better comparison between monomer and dimer samples, spectra were corrected for the difference in chlorophyll concentration. For further details, see "Materials and Methods." obtained ( Fig. 6A), which decayed rapidly upon dark adaptation at 273 K (Fig. 6B). The presence of a Q A semiquinone radical superimposed on signals from electron donors such as chlorophyll makes analysis of this complex spectrum difficult. However, a crude subtraction of the signal from the Q A semiquinone (chemically reduced; Fig. 6C), assuming an equal yield from chemical and photochemical reduction, produces Fig. 6D. Comparison with the characteristic spectrum of the tyrosine Y D radical from PSII membranes (Fig. 6E) shows that hyperfine peaks corresponding to those of the tyrosine radical are observed in Fig. 6D. This indicates that a tyrosine radical can be trapped in PSII dimers in addition to the narrower Chl ϩ or Car ϩ species, producing the mixed spectrum in Fig. 6D. Since the spectra of tyrosine Y Z and Y D radicals are very similar, we cannot distinguish between these species. DISCUSSION Chemical analyses has previously shown that the dimeric, but not the monomeric, form of the isolated CP47-RC complex contained plastoquinone-9 measured to be 1.5 Ϯ 0.3 molecules/RC (8). The fluorescence, thermoluminescence, and EPR data presented in this paper show that this plastoquinone can act as a secondary electron acceptor, indicating that the dimeric form of this complex is closer to being a functional PSII complex than the monomeric form. Indeed, the CP47-RC dimer contains more protein subunits than the monomer (see Introduction) and binds slightly more chlorophyll (21 Ϯ 2.5 compared with 18 Ϯ 1.5 molecules/RC (see accompanying paper (8)). However, the EPR data show that the signal due to photoinduced plastoquinone reduction in the dimer is not that of the iron-semiquinone typically found in more intact PSII systems. It seems highly likely that although the CP47-RC maintains Q A activity, the complex has lost its non-heme iron during isolation or that the iron-quinone interaction has been disturbed. The monomeric complex, which lacks Q A , is capable of primary charge separation and charge recombination between the radical pair P680 ϩ Pheo Ϫ as shown by the high fluorescence yield in weak measuring light and by the high yield of spin-polarized P680 triplet detected by EPR.
The capability of CP47-RC dimers to reduce Q A upon illumination in the absence of added exogenous electron donors implies the presence of active endogenous donor(s) in the complex. The dimers do not evolve oxygen and thus do not contain a functional manganese cluster that serves as a final electron donor for Q A reduction in active PSII centers. Treatment of the dimeric complexes with EDTA plus Tris, which completely removes manganese from PSII, had no effect on the variable fluorescence (not shown), excluding the possibility that some residual, nonfunctional manganese could act as electron donor. In the case of the D1-D2-cytochrome b 559 RC complex, the accessory chlorophylls and ␤-carotene can donate electrons to P680 ϩ (27,28). The observation that the fluorescence relaxation curves can be induced by repetitive flashes (Fig. 2) indicates that the endogenous donor rapidly recovers after its oxidation. This makes it very unlikely that the accessory  mers (B and D). Left, the g y (near 290 mT) and g z (near 220 mT) peaks of cytochrome b 559 are shown. The samples on the right were chemically reduced by dithionite treatment in the dark at 273 K before freezing. EPR conditions were as follows: microwave power (left), 4 milliwatts and (right), 1 microwatt; modulation amplitude (left), 2 mT and (right), 0.2 mT; temperature, 12 K. For further details, see "Materials and Methods." chlorophylls or ␤-carotene are the only endogenous electron donors, since they are quite stable in the oxidized state. The possibility that cytochrome b 559 acts as a donor is also unlikely, given that this heme protein is present in its low potential oxidized state in both forms of the complex. Since the EPR data indicate light-induced formation of a tyrosine radical (Fig. 6), which could be either Y Z ⅐ or Y D ⅐ , it is more likely that Tyr-Z, which has a much shorter lifetime in the oxidized radical form than Tyr-D, is the electron donor for Q A reduction. In that case, the relaxation of the flash-induced fluorescence yield increase would correspond to a reoxidation of Q A . via recombination with Y D ⅐ . This idea is in agreement with the 600-ms relaxation time of Y Z ⅐ and a similar reoxidation half-time of Q A . in Ca 2ϩdepleted PSII particles, which, like the CP47-RC dimers, lack a functional manganese cluster and have impaired Q A to Q B electron transfer (16). Thus, the detection of a tyrosine radical in the CP47-RC dimer indicates that this complex is functionally active on the donor side as well as the acceptor side. The perturbations in its functions compared with more intact PSII systems are probably linked to the absence of CP43 and the extrinsic proteins of the oxygen-evolving complex as well as to the loss of the non-heme iron from the acceptor side and the manganese cluster from the donor side. Furthermore, a distinct difference between the fluorescence properties of the dimeric CP47-RC and BBYs was the fast oxidation of Q A in BBYs compared with the CP47-RC dimer when the actinic light was turned off (Fig. 1, trace 1). This is expected, since BBYs maintain a plastoquinone pool and Q B activity. There has been discussion as to whether PSII normally exists as a monomer or dimer in vivo (1). Although the experiments presented in this paper were not conducted to enter into this debate, they do strongly suggest that the PSII complex in the membranes from which the CP47-RC complex was isolated are dimeric. The starting point for the isolation procedure are PSII-enriched membranes of the BBY type (29). These membranes are derived from the grana that contain the majority of functionally active PSII in higher plant chloroplasts (30). Previously, larger dimeric PSII complexes have been isolated from BBYs using mild detergent treatments. The largest to date is a 710-kDa supercore dimer complex containing LHC-II, CP29, and CP26 as well as CP43, CP47, and the RC proteins (31). It binds about 100 chlorophylls (Chl a and Chl b)/RC. It is relatively easy to strip away the LHC-II, CP29, and CP26 components to yield a dimeric core complex with a molecular mass in the region of 500 kDa and binding about 35 chlorophyll a molecules per RC (32). Two-dimensional crystals of this PSII core dimer complex have recently been analyzed by electron crystallography (to yield a three-dimensional map) and compared with a similar analyses of two-dimensional crystals of the CP47-RC complex (32). The results indicate that CP43 is located toward the outer edges of the core dimer. It is therefore reasonable to assume that the solubilization conditions used for the present study can strip away the peripheral CP43, leaving a dimeric form of the CP47-RC complex. In fact, as reported previously, the isolation procedure we have adopted produces mainly the dimeric form of the complex (8). The CP47-RC monomer is presumably derived from dissociation of the dimer, which leads to a loss of plastoquinone, some chloro-phyll, and the PsbL and PsbK proteins (8).
The observations presented in this paper show that there is no active Q A in the monomers, and its presence in the dimers excludes the possibility that dimers could be formed by association of monomeric CP47-RCs, since there seems to be no way to acquire a bound quinone at the Q A site during dimerization. The results also support the conclusion that PsbL plays a role in stabilizing the secondary electron transfer properties on the acceptor and donor sides of PSII (33,34).